INDEX

Explanations

words related to something harmful or deadly

oai_token-act-pair · gpt-3.5-turbo

terms associated with lethal force and weapons

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.hook_resid_pre

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

Data Type

torch.float32

Hook Point

blocks.1.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

rix

-0.83

 Transparency

-0.83

Rate

-0.77

 Stud

-0.77

ebook

-0.74

roo

-0.72

reads

-0.72

Hub

-0.72

Rail

-0.71

acular

-0.71

POSITIVE LOGITS

 lethal

1.32

 mosqu

1.02

otine

1.01

etheless

1.00

lethal

0.99

 injection

0.93

ciating

0.91

 assassins

0.90

espie

0.87

 assassin

0.87

Activations Density 0.006%

words related to something harmful or deadly

terms associated with lethal force and weapons

No Comments

No Known Activations